AI and Automation Offer Both Promise and Peril to the Hiring Process
CDW’s Core to Cloud Summit highlights the pros and cons of these technologies.
While CDW’s recent Core to Cloud Summit event in Las Vegas was ostensibly focused on cloud and data center technologies, there was a great deal of discussion among both presenters and participants about the people behind the machines. How are organizations tackling the difficulties of filling highly technical IT staff positions in a job market where there are too few candidates available with the requisite skill and experience? Is this tiny pool of talent a bug or a feature of the future IT job market? How will organizations address future staffing needs? Lots of questions and conversations, but few answers.
Applying Automation to Staff Shortfalls
Mike Palmer, chief product office at Druva, touched on the current IT labor market during his presentation “Why the Cloud Continues to be a Better Option for Data Protection.” One of his slides compared the top-paid IT certifications from 2008 to 2018, noting a clear shift from more generalized certs like Microsoft Certified Architect to AWS Certified Developer and Google Certified Professional Cloud Architect. IT operations continue to grow more complex and specialized, and organizations are finding it more difficult to recruit workers to meet their needs.
Automation has been one way that companies have tackled this scarcity of IT talent. Organizations have started leaning on new technologies such as artificial Intelligence (AI) and machine learning (ML) to handle responsibilities previously allocated to IT teams. Cybersecurity offers a good example of this shift. With experienced security specialists in high demand and not enough talent available, automated security solutions have been developed to fill the gap, automating time-intensive security activities like threat intelligence, configuration management, and incident detection and response.
The Problem with Machines
As AI and ML continue to evolve, they are set to shoulder a greater burden of IT responsibilities. The Core to Cloud event’s keynote by Judy Spitz, founding program manager at the Initiative for Women in Technology and Entrepreneurship in New York at Cornell Tech, touched on this advancing tide of AI that is blurring the boundaries between human-centered and machine-enabled work. She portrayed AI in a positive light, as an augmenter of human ability extending our productivity, rather than playing into the notion of AI replacing humans in their job roles.
But in many of its current iterations, AI and ML’s wondrous benefits are undercut by what Spitz termed “algorithmic bias.” People tend to view machines as inherently neutral and fair, while their fellow human beings are seen as biased. Applying algorithms to business processes holds the promise of quicker decision-making without the human subject bias. Unfortunately, the large data sets that underpin these technologies are produced by humans and are, therefore, flawed from the start by their inherent biases.
AI and ML have been applied to the hiring process to speed up and automate what’s a very labor-intensive process, with mixed results. Amazon canceled a project that was using AI to vet job applications because the application was consistently downgrading female candidates.
Growing Workplace Diversity
The issue of algorithmic bias comes at a time when companies are looking to grow a more diverse workforce. More companies are beginning to connect the dots linking diversity in the workplace to a stronger bottom line, greater outputs from their teams and improved recruitment opportunities. Working within teams that have diversity — whether it is gender, race, sexual identity, age, experience or background — requires staff to have soft “people” skills that are difficult for computer applications to evaluate. Applying AI and ML to hiring doesn’t appear to offer much assistance in efforts to grow diversity in the workplace.
These technologies certainly have their place in IT operations, but like so many new technologies, companies need to allow the hype to dissipate so they can see more clearly where to plug them in. Human being will always be needed, especially for the subjective, unique — diverse — perspective that each of us brings to everyday business challenges. The machines will never be able to replace that.